2# NEW BOOKS and MULTIMEDIA

li

/
this book by providing the (C-like) pseu- cally the cryptographic techniques deployment. Part of the problem is
docode of what essentially happ’ens include DES, 3DES, IDEA, SAFER, purely the syntactical aspect of such
behind the scenes when certain c m-
mands are invoked (we are warned not
t
to expect that the pseudocode is a qne-
to-one reflection of the actual Ci co
Blowfish, CAST-128, RC2, RC4, RC5,
RC6, AES, RSA, Diffie-Hellman, EIGa-
mal, DSS, ECC, digital envelopes, key
protection, and pseudorandom
protocols (performance being the other
one) that allow, for example, SCSI com-
mands to be conveyed via IP. To this
extent, t h e book presents the SCSI
source code, but rather a summariz d
idea of the control flow). That beidg
l+ sequences, as well as certain legal issues.
The reader should be warned that the
Architecture Model (SAM-2), and SCSI
architecture in general ( C h a p t e r 4)
said, understanding the internals boil
down to understanding the pseudocode, P presentation of cryptographic techniques
I .

is severely condensed (the topics listed

which is key to understanding the rest.
A sequence of chapters ( 5 , 6, and 7 )
possibly alienating some readers n t
keen on reading code. Nevertheless, one sl’ above take less than 28 pages of the
book); if interested in the properties and
provide a review of IP, UDP, and TCP
for those unfamiliar with networking.
can approach the pseudocode on an hs limitations of a scheme, instead of just its T h e rest of the book centers around
needed basis, when the behavior of c Im- application, the reader is advised to seek IETF’s Fibre Channel over IP (FCIP),
mands is not exactly as intuitive as Xne other sources of information. From the I n t e r n e t Fibre Channel Protocol
might expect. The reader is assumed to remaining parts, the second is devoted to (iFCP), and I n t e r n e t SCSI (iSCSI).
be familiar with networking, but a short firewalls: packet filtering, SOCKS, appli- These protocols are described in detail,
review of IP addressing and its relation cation-level gateways, configuration of and they correspond to, respectively, FC
to routing is nevertheless provided in the firewalls, and in gcneral the architectural extensions to connect remote F C SANS
introduction. A chapter describes the elements of a potentially secured net- via IP (FCIP), mapping between IP and
routing information organization and work. The communications part pays FC end nodes (iFCP), and replacement
management inside Cisco routers, fol- attention to a handful of specific proto- of FC and devices with native IP storage
lowed by a chapter that describes the IP cols at different layers of the protocol devices (iSCSI). Another important fac-
forwarding process options present in stack, explaining their principles of oper- tor of IP SANs, especially in larger
Cisco routers: fast, optimum, distributed, ation. These include MS-PPTP, IPsec, installations, is their management; this is
NetFlow, and Cisco Express forwarding. SKIP, I D E , SSL, TLS, Kerberos, covered in Chapter 9 through presenta-
After the internal mechanisms and for- SESAME, PGP, and SWIME. The book tion of the Internet Storage Name Server
warding options come the chapters on ends with a discussion of PKI (certifi- (iSNS) protocol. To complete the pre-
actual routing protocols: static routing, cates and certificate authorities), and a sentation, security and quality of service
dynamic routing, distance vector proto- short overview of electronic commerce, of S A N S are presented, even though they
cols (RIP and IGRP), and link state pro- risk management policies and principles. are not as closely related to SANs alone
tocols (OSPF and IS-IS). A separate Overall, the book is written for an audi- per se, but rather to IP security and QoS.
chapter describes Enhanced IGRP, ence without necessarily a significant Arguments to convince the reader of the
Cisco’s proprietary protocol. The style is technical background who wish to quick- value of the application potential of
geared toward answering specific ques- ly rcview a broad introduction to the SANs can be found in Chapter 13, from
tions as to why things work in a particu- available techniques and protocols of obvious choices (backup) to perfor-
lar way, instead of just how. Consistent network security. mance-driven applications (server clus-
with the book’s scope, each chapter tering and massive data distribution).
includes several examples and is followed If‘ SANS,A Guide to iSCSI/ Appendices provide information about
by a Q-and-A section of the most fre- the Storage Networking Industry Associ- ‘
quently asked questions. iFCP, and FCIP, Protoco/s for ation (SNIA) and IP storage (and rela’t-
Storage Area Networks ed) vendors.
Internet and lntranet S e w r i b Tom Clark, 2002, Addison-Wesley,
2nd Edition ISBN 0-201-75277-8, 288 pages, soft- java.rmi The Remote Method
cover. Invocation Guide
Rolf Opplinger, 2002, Artech House,
ISBN 1-58053-166-0, 403 pages, The objective of storage area networks Esmond Pitt and Kathleen McNiff,
hardcover. (SANs) is to produce logical storage 2001, Addison-Wesley, ISBN 0-201-
units that are in fact scalable collections 70043-3, 2 8 4 pages, softcover.
Rolf Opplinger’s book is an attempt to of storage devices over a network.
present security in networks (intra- and While storage devices are well appreci- The idea of distributed programming
internets) in a unified fashion. For a ated as scalable in terms of capacity at a across heterogeneous CPUs, OSs, lan-
long part of the book, networking and nearly exponential rate, the extent to guages, and data formats has galva-
cryptography are presented separately, which a specific network technology is nized researchers and software
coming together in the second half to be uscd and its shortcomings, in developers alike, producing, for exam-
(“Communications Security”). The con- terms of scalability, are still an issue for ple, RPC and CORBA. The emergence
tents cover not just cryptographic research and development. One prevail- of Java as a universal programming
aspects of security but also general-pur- ing school of thought advocates the use language led, naturally, to its use as a
pose security mechanisms that can be of IP as the network technology of step toward simplifying distributed sys-
used to enhance security, such as packet choice. The deployment of SANs over tems programming. The bulk of func-
filtering and firewalling. The first part IP requires the solution to a plethora of tionality in s u p p o r t of distributed
lays the foundations, including terminol- problems because of the need of a lega- applications under Java is known as the
ogy, basics of TCPIIP, types and nature cy general-purpose internetworking pro- R e m o t e Method Invocation ( R M I )
of attacks, the OS1 security architecture, tocol (IP) to efficiently support API. RMI is already quite popular in
and a brief review of cryptographic relatively recent protocols, of special- the development of applications, but
techniques followed by authentication ized n a t u r e (SCSI, Fibre Channel), the range of information necessary
and key distribution techniques. Specifi- some not even intended for WAN (which goes beyond the Java RMI doc-
~

IEEE Network March/April2002 5

c1 NEW BOOKS and MULTIMEDIA

umentation) to fully exploit its power is in the periphery of RMI proper, but addition of LAN operating systems. A
not widely available. The book by Pitt likely to be encountered when deploy- single chapter describes TCP/IP, from
and McNiff hopes t o fill this gap as ing elaborate or specialized applica- addressing to IP to essentials of routing
well as t o raise b o t h awareness of tions. and transport layer protocols (TCP and
RMI's potential and to link the con- UDP), ending with applications (SMTP,
cepts of distributed programming t o Networks for Computer Scien- FTP, POP3, Telnet, DNS) as well as
corresponding constructs in R M I , dialup data link protocols (SLIP, PPP).
exposing, in the pi-ocess, the internals tists and Engineers What is surprising is the relatively
of RMI and its des,ign. The audience is YGlu Zheng and Shakil Akhtar, 2002, sparse information on HTTP despite a
expected to be reasonably knowledge- Oxford University Press, ISBN 0-19- section devoted to Web applications.
able with Java as well as familiar with 5 1 1398-5, 575 pages, hardcover. The next theme discusses access and
0-0concepts (inheritance and poly- high-speed technologies. This is really a
morphism). In particular, the reader is Among available undergraduate net- chapter on both the last mile, as well as
assumed to be aware of the networking works texts, t h e o n e by Zhcng and on LAN and WAN protocols, namely
classes Socket and Serversocket. The Akhtar follows a particularly interesting ISDN, DSL, SMDS, Frame Relay,
newcomer t o R M I will find the first structure. First, two example applica- FDDI, High Speed Ethernet, FDDI and
nine chapters the most helpful to start tions introduce corresponding architec- CDDI, ATM, SONET, and D W D M .
writing applications. These chapters tures: voice 'communication for The switching and virtual LAN chapter
include, apart from introduction and telephone networks on one hand, client- spans from hubs to LAN switches, intro-
concepts, the semantics of RMI, serial- server applications for packet-switched ducing VLANs and eventually A T M
ization, remote interfaces, the R M I networks on the other. Mechanisms are switches, but lacking in terms of
registry, basic client and server pro- presented to illustrate how such net- describing bridging protocols. Network
gramming, basic security, and tcch- works operate. The examples lead to performance presents analytical (basic
niques/support f o r mobile code more detailed information, such as the queuing), simulation, and network traf-
execution. T h e p r o g r a m m e r m o r e distinction along the lines of topology fic monitoring. The network manage-
familiar with distributed programming and the need for standards. The second ment chapter covers b o t h t h e usual
and elements of RMI finds better use part deals with the principles of data suspects (SNMP and R M O N ) and
for the following three chapters that communications at the level of signals telecommunications management net-
provide detailed information on activa- and modulation but is, oddly, comple- work (TMN). The coverage of cryptog-
tion (essential to provide the illusion of mented by a section on videoconferenc- raphy and security is fairly
persistent reference to an object in a ing (applications and standards). The comprehensive (secret and public key
distributed environment), socket facto- OS1 reference model is next, but once crypto, DES and AES, MD5, PGP, PKI,
ries, and example architectures for more the chapter includes, oddly firewall architectures and types, Ker-
agents programming as an example of enough (since a separate chapter on the heros, VPNs, smart cards, etc.). Con-
the design patterns facilitated by the topic exists), an introduction to network cluding is a chapter o n network
RMI programming paradigm. The last performance. T h e OS1 model is fol- programming, with a typical introduc-
part of the book presents alternatives lowed by LAN technologies (the majori- tion to sockets (and Winsock) program-
and extensions t o the standard R M I ty of which is a description of the IEEE ming plus low-level programming over
implementation (JNDI, naming ser- 802 standards), including hardware serial or parallel ports, as well as basics
vices, Jini, CORBA/IIOP, SSL): topics aspects, with the somewhat misplaced of RPC and Java programming.

0 IEEE 802 PERSPECTIVES/Edited by Paul Nikolich

The IEEE 802 L.ocal and Metropoli- vide the reader with highlights of IEEE p.nikolich@ieee, IEEE 802 chair.
tan Area Network (LAN/MAN) Stan- 802 activities to enable better dissemina- As a final note, Jim Carlo retired
dards Committee is an international tion of these standards into marketable from the 802 Chair position in Novem-
standards developing organization con- products as well as seek new ideas to be ber 2001. I wish to thank Jim for the ._
sisting of more than 1000 individual net- brought into the IEEE 802 arena. leadership and mentorship he has pro-
working experts located worldwide. The This perspective will provide a brief vided to IEEE 802 since he took over as
major industry standards now in devel- overview of the I E E E 802 standards Chair in 1996.
opment by I E E E 802 a r e E t h e r n e t development process. A Web page Paul Nikolich
(CSMA/CD), Wireless Local Area Net- where additional information can be
works (WLAN), Wireless Personal Area obtained on IEEE 802 activities, meet- Biography
Networks (WPAN), Broadband Wire- ing dates, and organization is located at PAUL NIKOLICH
wos elected to the IEEE802 Chair in
less Access (BWA), Resilient Packet http://www.ieee802.org. November 2001 after serving as IEEE 802 Vice
Ring (RPR), and Bridging/Architecture Subsequent columns will contain Chair since 1996. He has led product, technology,
that provide the basis for enterprise net- highlights of specific standards in devel- and standaids development activities at several
working. There are about 50 IEEE 802 opment. Some of the exciting topics in companies since the early 1980s. Most recently he
LANJMAN standards and five major IEEE 802 are 10 Gigabit Ethernet and was co-founder of Broadband Access Systems, Inc.,
ISO/IEC/JTCl 8802 series equivalent how Ethernet will migrate outward from a supplier of carrier class networking equipment to
standards covering the broad spectrum the enterprise to the first mile, powered cable television system operators. He currently i s on
of LAN/MAN standards. Ethernet, and the growing diversity of independent consultant working closely with YAS
These perspectives have appeared in wireless communication networks. Broadband Ventures, LLC on investing in early stage
some issues of I E E E Network since Please send suggestions, criticisms, and startup companies
September 1997 and are aimed to pro- requests for future article topics t o

6 IEEE Network March/April 2002